Prostate cancer (PC) is the commonest cancer in men and is highly heterogeneous; despite this, molecular stratification to guide treatment is not standard of care. Defects in DNA repair genes in sporadic PC present a therapeutic opportunity. Preclinical studies indicate that cell lines defective in genes involved in homologous recombination, including BRCA1/2, ATM, ATR, CHEK2 and RAD51, are sensitive to PARP inhibitors (PARPi). We have now shown that PARPi have antitumour activity in sporadic PC in clinical trials of niraparib (Sandhu et al, 2013) and olaparib (unpublished). AIMS I. To determine the frequency of DNA repair defects in sporadic PC and the impact of these on prognosis and response to endocrine treatment. II. To determine the prevalence of DNA repair defects in patients responding or resistant to PARPi. III. To evaluate the functional relevance, in preclinical models, of genomic aberrations identified in these tumor biopsies studied in Aims I and II. IV. To clinically qualify a biomarker suite that will serve as a predictive biomarker of sensitivity to PARPi in sporadic CRPC. METHODOLOGY 1. Next generation sequencing (NGS) of CRPC samples using a customized panel to identify DNA repair defects. These data will be correlated with clinical data. 2. Use of whole genome sequencing data of sporadic CRPC responding or resistant to PARPi to identify defects in DNA repair that associate with response. 3. Knockout and knock-in functional studies of identified genes, in in vitro studies, to determine their impact on PARPi sensitivity and DNA repair. 4. Customized, high throughput, targeted and high coverage NGS (>1000 fold coverage) using the MiSeq Illumina platform of all the identified genes to prospectively select patients for planned large randomized studies. IMPLICATIONS: It is envisioned that these studies will generate the first biomarker suite to guide systemic treatment selection for prostate cancer.